High-precision surface defect detection in manufacturing often relies on laser triangulation profilometric sensors for detailed surface measurements, providing detailed and accurate surface measurements over a line. Accurate motion between the sensor and workpiece, usually managed by robotic systems, is critical for maintaining optimal distance and orientation. This paper introduces a novel Reinforcement Learning (RL) approach to optimize inspection trajectories for profilometric sensors based on the boustrophedon scanning method. The RL model dynamically adjusts sensor position and tilt to ensure consistent profile distribution and high-quality scanning. We use a simulated environment replicating real-world conditions, including sensor noise and surface irregularities, to plan trajectories offline using CAD models. Key contributions include designing a state space, action space, and reward function tailored for profilometric sensor inspection. The Proximal Policy Optimization (PPO) algorithm trains the RL agent to optimize these trajectories effectively. Validation involves testing the model on various parts in simulation and performing real-world inspection with a UR3e robotic arm, demonstrating the approach’s practicality and effectiveness.

In the manufacturing industry, inspection systems play a crucial role in ensuring product quality. High-resolution profilometric sensors have become increasingly popular for inspection due to their ability to provide detailed surface information. However, the development and testing of inspection systems can be costly and time-consuming. This paper presents the development of a simulation of an inspection system using a high-resolution profilometric sensor. A geometrical and noise model is proposed to simulate the readings of any actual profilometric sensor. The model replicates the sensor's movement on the CAD model of the inspected part. The model incorporates the physical properties of the sensor and combines noise sources from sensor uncertainty and speckle noise induced by the roughness of the material. Our contribution lies in noise modeling. This work proposes a combination of Perlin noise to simulate the speckle noise and Gaussian noise for the uncertainty-related noise. Perlin noise is generated based on the surface roughness parameters of the inspected part. The accuracy of the simulation system is evaluated by comparing the simulated scans with real scans. The results highlight the ability to simulate real scans of different parts, using commercial sensor specifications and the CAD model of the inspected part.

The problem of restoration of a high-resolution image from several degraded versions of the same scene (deconvolution) has been receiving attention in the last years in fields such as optics and computer vision. Deconvolution methods are usually based on sets of images taken with small (sub-pixel) displacements or slightly different focus. Techniques based on sets of images obtained with different point-spread-functions (PSFs) engineered by an optical system are less popular and mostly restricted to microscopic systems, where a spot of light is projected onto the sample under investigation, which is then scanned point-by-point. In this paper, we use the effect of conical diffraction to shape the PSFs in a full-field macroscopic imaging system. We describe a series of simulations and real experiments that help to evaluate the possibilities of the system, showing the enhancement in image contrast even at frequencies that are strongly filtered by the lens transfer function or when sampling near the Nyquist frequency. Although results are preliminary and there is room to optimize the prototype, the idea shows promise to overcome the limitations of the image sensor technology in many fields, such as forensics, medical, satellite, or scientific imaging. (C) 2017 Elsevier Ltd. All rights reserved.

This article presents, by means of computational simulation tools, a full analysis and design of an Interferometric Fiber-Optic Gyroscope (IFOG) prototype based on a closed-loop configuration with sinusoidal bias phase- modulation. The complete design of the different blocks, optical and electronic, is presented, including some novelties as the sinusoidal bias phase-modulation and the use of an integrator to generate the serrodyne phase-modulation signal. The paper includes detailed calculation of most parameter values, and the plots of the resulting signals obtained from simulation tools. The design is focused in the use of a standard single-mode optical fiber, allowing a cost competitive implementation compared to commercial IFOG, at the expense of reduced sensitivity. The design contains an IFOG model that accomplishes tactical and industrial grade applications (sensitivity ≤ 0.055 °/h). This design presents two important properties: (1) an optical subsystem with advanced conception: depolarization of the optical wave by means of Lyot depolarizers, which allows to use a sensing coil made by standard optical fiber, instead by polarization maintaining fiber, which supposes consequent cost savings and (2) a novel and simple electronic design that incorporates a linear analog integrator with reset in feedback chain, this integrator generating a serrodyne voltage-wave to apply to Phase-Modulator (PM), so that it will be obtained the interferometric phase cancellation. This particular feedback design with sawtooth-wave generated signal for a closed-loop configuration with sinusoidal bias phase modulation has not been reported till now in the scientific literature and supposes a considerable simplification with regard to previous designs based on similar configurations. The sensing coil consists of an 8 cm average diameter spool that contains 300 m of standard single-mode optical-fiber (SMF-28 type) realized by quadrupolar winding. The working wavelength will be 1310 nm. The theoretical calculated values of threshold sensitivity and dynamic range for this prototype are 0.052 °/h and 101.38 dB (from ±1.164 × 10−5 °/s up to ±78.19 °/s), respectively. The Scale-Factor (SF) non-linearity for this model is 5.404% relative to full scale, this value being obtained from data simulation results.

Conical diffraction is an interesting phenomenon that occurs in biaxial crystals. Although the existence of this conical diffraction has been understood for a very long time, it was not until recently that the first applications started to appear, and most of them focus on the manipulation of laser beams. However, the potential use of conical diffraction for all-optical processing in full-scene imaging devices was never addressed. In this article we present a prototype of a general-purpose optical device in order to investigate the applicability of conical diffraction to full-scene imaging. We also show that conical diffraction can obtain the derivative of an image with high resolution. This is the first proof of the theoretical capabilities of conical diffraction in full scenes, and it is just one of its many applications in the field of machine vision.

Characterizing waviness in sheet metal is a key process for quality control in many industries, such as automotive and home appliance manufacturing. However, there is still no known technique able to work in an automated in-floor inspection system. The literature describes many techniques developed in the last three decades, but most of them are either slow, only able to work in laboratory conditions, need very short (unsafe) working distances, or are only able to estimate certain waviness parameters. In this article we propose the use of a lateral shearing interferometric profilometer, which is able to obtain a 19 mm profile in a single acquisition, with sub-micron precision, in an uncontrolled environment, and from a working distance greater than 90 mm. This system allows direct measurement of all needed waviness parameters even with objects in movement. We describe a series of experiments over several samples of steel plates to validate the sensor and the processing method, and the results are in close agreement with those obtained with a contact stylus device. The sensor is an ideal candidate for on-line or in-machine fast automatic waviness assessment, reducing delays and costs in many metalworking processes.

We describe the automated application of an area based registration method to the surface inspection of steel industry products as a tool to solve an intermediate mosaicing problem. The main problem of area based methods is that there is high probability that the results of a matching process will be incorrect if a region of interest without any relevant detail is used. The selection of a region of interest with relevant content continues to be a problem nowadays. We propose a method to select a salient area when using a zero mean normalised cross correlation metric and a block as a region of interest. The selection of the size and the position of the block is focused on ensuring a smooth unimodal similarity surface around the maximum similitude point. Experiments show a correlation between the surface kurtosis of the block autocovariance and the same coefficient measured over the correlation surface around the maximum similitude point for the three different steel products analysed. We check that the maximum correlation value is reached abruptly, in a small range of pixels around the maximum similitude point, in correlation surfaces obtained from blocks containing non-relevant information. On the other hand, salient blocks usually lead to unimodal smooth similarity surfaces with small sensitivity to noise in contrast with the ones obtained from non-remarkable blocks. Also, the method proposed allows the application of fast search algorithms based on the unimodality of the correlation surface, obtaining high computational time reduction in comparison with full search strategies using fast normalised cross correlation algorithms.

In this paper we describe an interferometric profilometer which provides absolute distance measurements of rough surfaces in the submicrometric range, suitable for in situ application. The apparatus is based on lateral shearing interferometry and works as a range finder. Its main advantages are high acquisition speed, single-shot profile measurements even for moving objects, large field of view and depth of field, full common-path collinear setup, and in situ applicability for automatic measurements. Along with the theoretical study we provide a reconstruction method which makes use of the slope information contained in the fringe pattern to increase resolution, and experimental results that verify the applicability of the system. Our prototype obtains profiles with an aperture of 19mm from a working distance larger than 90mm, a depth of field over 4mm, lateral resolution of 10μm, and depth resolution well below 1μm, at a speed of 30 profiles per second. However, the system can be easily scaled to meet specific application requirements.

In this paper, we propose a method to compare and visualize spectrograms in a low dimensional space using manifold learning. This approach is divided in two steps: a data processing and dimensionality reduction stage and a feature extraction and a visualization stage. The procedure is applied on different types of data from a hot rolling process, with the aim to detect chatter. Results obtained suggest future developments and applications in hot rolling and other industrial processes.

A common-path interferometric profilometer using a Savart plate as a lateral shearer has been successfully tested under harsh environmental conditions to measure the shape of a surface, detecting defects and characterizing surface properties. The whole profile is obtained from a single image and its depth sensitivity is easily scalable, making this technique suitable for many different applications. Although this system has been successfully used for surface inspection and defect detection, some behaviors cannot be explained by the usual simple model for fringe formation, which, amongst other things, considers normal incidence of the incoming rays into the Savart plate. These deviations from the ideal case are more noticeable for high resolutions from short distances. This paper studies the formation of the fringe pattern, which is crucial for understanding the behavior of the system and proper calibration.

We propose a common-path two-wavelength interferometric system based on a single optical element, a Savart plate, that is able to obtain single-shot profile measurements with submicron precision from safe working distances (beyond 100 mm). These characteristics make this sensor ideal for surface inspection in on-line applications. For the illumination branch, two lasers with close wavelengths are combined and then passed through a rotating holographic diffuser for drastic speckle reduction. In the acquisition branch, the interferometric signals of both wavelengths are captured simultaneously by a camera, and their phase signals are combined to extend the measurement range.

Conoscopic Holography proved to be a very adequate solution for in-situ optical measurement in industrial inspection and quality control systems, offering high-precision with a wide range of standoff distances, while being quite insensitive to the harsh environmental conditions often encountered in industry, as it is a common-path technique.With the aim of extending their applicability, we have already addressed, with good results, several issues that improve sensors based on this technology which include: the use of phase information to obtain one-shot profile measurements at frame rate with higher precision; new signal processing techniques; and speckle reduction to diminish measurement errors. However, the undesirable effect of using the phase information is that it reduces the maximum steep that can be measured without ambiguity, which becomes an issue when working with high precisions.In this article we present our ongoing work towards using the concepts of multiple-wavelength interferometry to extend the measurement range, something that, to our knowledge, has not been done for this technology before.

On-line non-contact surface inspection with high precision is still an open problem. Laser triangulation techniques are the most common solution for this kind of systems, but there exist fundamental limitations to their applicability when high precisions, long standoffs or large apertures are needed, and when there are difficult operating conditions. Other methods are, in general, not applicable in hostile environments or inadequate for on-line measurement. In this paper we review the latest research in Conoscopic Holography, an interferometric technique that has been applied successfully in this kind of applications, ranging from submicrometric roughness measurements, to long standoff sensors for surface defect detection in steel at high temperatures.

Early surface defects inspection in hot steel products is a difficult task, but can help to reduce significantly production costs. This is the case of steel slabs when they are produced in the continuous casting line. Conoscopic holography phase-based long stand-off profilometers have shown to be a great tool for this kind of inspection, and a breakthrough system based on them is being used for more than 2 years in production conditions with high reliability and economical impact. This paper presents the results of this system and the challenges it has overcome: hot material up to 900°C, dust, scale over the inspected surface.

On-line noncontact surface inspection with high precision is still an open problem. Usual methods are, in general, not applicable in hostile environments or not adequate for on-line measurement, as they are either slow in nature or need to work from very short, unsafe distances, providing small depths of field and apertures. The ongoing work toward the development of a noncontact optical profile measuring sensor that could be used for submicron measurements in on-line applications is presented here. Our approach is based on conoscopic holography and triangulation, and uses a very simple method for removing speckle noise, which is key for obtaining high precisions from safe distances (several centimeters). (C) 2008 Society of Photo-Optical Instrumentation Engineers.

One of the more challenging applications of optical metrology is real- time dimensional control and surface inspection in industrial applications, where strong requirements of cost, speed of operation, ease of setup and applicability in adverse environments greatly limit the number of applicable technologies. The design and uses of an optic profilometer, based on conoscopic holography, have been reported previously, but there are still some drawbacks that should be addressed. One of the most important is signal processing, which is a relatively expensive process and limits the acquisition rate at no more than 70 profiles per s. We present a new approach to the signal processing problem, deriving that the phase information contained in one fringe pattern, which corresponds to one profile, can be considered as a combination of multiple one-dimensional (1-D) patterns that carry the same information but for a phase difference in the carrier wave, making it possible to apply efficient phase-shifting interferometry (PSI) techniques and resulting in a reduction of more than two orders of magnitude in computational needs. (C) 2007 Society of Photo-Optical Instrumentation Engineers.

One of the more challenging applications of optical metrology is real-time dimensional control and surface inspection in industrial applications, where strong requirements of cost, speed of operation, ease of setup and applicability in adverse environments, greatly limit the number of applicable technologies.An optic profilometer, based on Conoscopic holography, has been designed specifically for this purpose and used in several on-line inspection systems. This device is able to obtain a distance profile of a target in a single shot; works at long distance standoff (700-1200 mm) and still keeps good resolution (under 0.2 mm) with a very easy and reliable setup.However, there are still some drawbacks that should be addressed. The first one is the signal processing, which is a relatively expensive process and limits the acquisition rate at no more than 70 profiles per second. The second one is speckle noise, which is an inherent problem in systems that use coherent-light illumination and triangulation, and therefor could be extrapolated to many other optical inspection systems.This paper shows the current lines of research to solve both problems and presents some initial, yet very interesting, results. These improvements can be applicable to other ranges of devices using this technology in adverse environments, for roughness and vibration measurement or surface defects detection.

On-line non-contact roughness metrology is still an open problem. Usual methods involve either contact (stylus-type devices) or perform indirect evaluations of some roughness parameters, such as Ra, with light scattering techniques or speckle measurement (among the most common optical techniques), inductance (only for magnetic materials) or ultrasound methods. However, a generic method able to obtain every roughness parameter (what means recording the real distance profile), able to work with a variety of surface types, and able to be installed in production lines is still to be developed. In this article, the ongoing work towards the construction of a non-contact optical profile measuring sensor that could be used for roughness measurements is presented. Our approach is based on Conoscopic holography, a common-path interferometric technique which is a good candidate for industrial applications. Current research effort is focused in enhancing accuracy in these systems, by both reducing the coherence of the illuminating source (laser) and changing the hardware and software setup, with the aim of building a sensor able to capture a profile of an object's surface in a single shot with high precision from a relatively long standoff (several cm).

One of the more challenging applications of optical metrology is real-time dimensional control and surface inspection in industrial applications, where strong requirements of cost, setup and applicability in adverse environments, greatly limit the number of applicable technologies. This paper shows an optic profilometer developed specifically for this purpose. This device, based on Conoscopic holography, is able to obtain a distance profile of a target in a single-axis scan; works from long distances and still keeps good resolution with a very easy and reliable setup. The first part of the paper introduces the working principles of Conoscopic holography and shows the sensor set-up. Necessary algorithms for obtaining the distance information are presented and the whole process is illustrated with real captures of test objects. The second part focuses on a real example of this technology applied in an on-line inspection system in steel continuous casting funded by the European Committee for Steel and Carbon, and which is currently working in Aceralia LDA steelmaking factory in Asturias (Spain). The system is placed in the process line and performs on-line detection of surface defects over hot steel slabs from a distance of 1200 mm. 100% of the production can be inspected without interfering with the process and without adding any delay.